https://github.com/mariantaragel/format-bench

Benchmark of data formats
https://github.com/mariantaragel/format-bench

benchamark data-formats python

Last synced: 26 days ago
JSON representation

Benchmark of data formats

• Host: GitHub
• URL: https://github.com/mariantaragel/format-bench
• Owner: mariantaragel
• Created: 2024-03-29T16:45:27.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2024-04-29T11:55:21.000Z (8 months ago)
• Last Synced: 2024-04-30T12:43:12.678Z (8 months ago)
• Topics: benchamark, data-formats, python
• Language: Python
• Homepage:
• Size: 40 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# FormatBench

FormatBench is a Python benchmark of data formats. This project aims to evaluate different data formats for storing tabular and image data.

Check out also: [ASV FormatBench](https://github.com/mariantaragel/asv-format-bench)

## Usage

```

python3 main.py (--tabular|--compression|--image) [--webface ] [--report]

```

`--tabular` - run tabular benchmark suite


`--compression` - run compression benchmark suite


`--image` - run image benchmark suite


`--webface` - run benchmarks with the Webface10M dataset; `` is a path to the Webface10M dataset


`--report` - generate report from the benchmark results


## Examples of usage

Run tabular benchmarks: `python3 main.py --tabular`

Run image benchmarks and create report: `python3 main.py --image --report`

Run compression benchmarks with the Webface10M dataset: `python3 main.py --compression --webface ~/synthetic_webface10M.h5`

Run tabular benchmarks with the Webface10M dataset and create report: `python3 main.py --tabular --webface ~/synthetic_webface10M.h5 --report`

## Related publication

TARAGEĽ, Marián. *Column-oriented and Image Data Format Benchmarks*. Brno, 2024. Bachelor’s thesis. Brno University of Technology, Faculty of Information Technology. Supervisor Ing. Jakub Špaňhel

## Acknowledgements

I would like to convey my gratitude to Ing. Jakub Špaňhel for his supervision. I also express my thanks to my consultant Ing. Petr Chmelař. Both of them provided me with support and advice during the work on this thesis. Last but not least, I would like to thank the external submitter, the Innovatrics company, for their professional help.